Automating the booking of deliveries, installations and technical visits means having an agent read each written customer request, apply the company's rules and book the service in the right calendar, with the confirmation sent back to the customer. The team no longer manages the back-and-forth behind every booking and only handles the requests that fall outside the defined rules.
In many small and medium-sized companies, bookings still depend on one person who reads the email, opens the field team's calendar, looks for a free day and replies to the customer. When the proposed day does not suit, the cycle starts again. The work is simple, but it happens every day and takes up the time of whoever should be coordinating operations.
What is a bookings agent?
A bookings agent is an artificial intelligence agent that interprets booking requests received in writing and proposes the next step, within the rules defined by the team. The booking itself is carried out by the platform, in the calendar that matches the requested service.
The agent handles requests for deliveries, installations and technical visits, as well as changes and cancellations to existing bookings. The same approach applies in other sectors and works in exactly the same way for appointments at a clinic or services at a garage.
The difference from an online booking page lies in where the process starts. On a booking page, the customer picks a slot in a calendar. With a bookings agent, the customer writes the way they always have, for instance “could you come and fit the shelving on Thursday morning?”, and the system interprets the request and fits it into the calendar. For organisations that mostly sell to other businesses, whose customers make requests by email, that difference matters.
Why does booking a delivery or a visit take so much effort?
Each booking looks like a small task, but it takes several rounds of emails and phone calls. The same steps come back with almost every request.
- Reading the request and working out which service is needed, for which address and for which order or piece of equipment.
- Confirming who the customer is and whether the request relates to something the company has sold or agreed to provide.
- Opening the right team's calendar and finding a free day that respects the customer's preference.
- Replying, waiting for confirmation and, quite often, starting over because the day no longer works.
- Letting the field team know and reminding the customer as the date approaches.
The cost is not only the time spent. A booking placed in the wrong calendar, an address copied incorrectly or a request forgotten in the inbox ends in a wasted trip and an unhappy customer.
How does automated booking work, step by step?
With EngiAIFlow, Engibots' intelligent process automation platform, the bookings process follows these steps.
- Intake. The request arrives by email or from another system through the API, and each request becomes a case.
- Reading. The agent interprets the request with artificial intelligence and extracts the type of service, the customer, the address and the preferred date. Each value keeps the sentence of the email it was read from.
- Rules and availability. The platform applies the company's rules to every booking and works out calendar availability, without AI.
- Booking and confirmation. The service is booked in the right calendar and the customer receives a reply through the same channel, with the confirmed booking and the calendar invitation attached.
- Team decision. Requests that fall outside the rules reach a person together with what has already been done and the decision still needed.
Changes and cancellations requested by the customer follow the same path. Automatic notices go out when a booking is confirmed, changed or cancelled, and as a reminder the day before.
Imagine, as an example, a company that sells and installs office furniture and has three installation teams. A customer writes asking for delivery and installation of an order the following week, preferably in the morning. The agent identifies the customer, the order and the address. The platform looks for a free morning that week in the teams' calendars and applies the company rule that an installation takes a full morning. The booking is made and the customer receives the confirmation with the invitation. If the address is outside the area the company covers, the request does not go through on its own and reaches the planning team with the information already prepared.
What should AI do, and what should the platform do?
Artificial intelligence is useful for understanding requests written in many different ways. Knowing whether a slot is free, on the other hand, is a calculation, and a calculation should always give the same result.
| Task | Who does it |
|---|---|
| Reading the request and extracting the service, the customer, the address and the requested date | Artificial intelligence, with the source sentence for each value |
| Confirming that the sentence for each value exists in the request | The platform |
| Working out calendar availability | The platform, without AI |
| Applying the company's rules | The platform, with the rules set up by the team |
| Deciding on a request outside the rules | A person on the team |
In EngiAIFlow, every value AI extracts from an email comes with the sentence it was read from, and the platform confirms that sentence exists. Without the sentence, the value is routed to a person. That way, a date the customer never wrote does not make its way into the calendar on its own.
The company's rules are managed in versioned sets, and each case's history shows which rule was applied. Rules are written in plain language, such as “before replying to the customer, ask for approval”.
How does the team stay in control of bookings?
A wrong booking has consequences outside the company, because it ties up a team and a customer. Control therefore has to be designed in from day one.
- Progressive autonomy. There are three levels, Apprentice, Supervised and Autonomous, and an agent described in words starts as Apprentice. The team raises the level at its own pace, as it gains confidence in the bookings made.
- Corrections that teach. The team's corrections guide the cases that follow. When the team links a sender to a customer, later messages from that sender arrive already identified.
- No repeated writes. A write to a connected system is never repeated automatically. If the process is interrupted, a person decides whether the write was already made.
- Case history. The history records what was done at every step and every human decision, with the change and the reason.
- Stop and pause. Whoever administers the installation can stop, with a single command, all external actions and every incoming channel, and the reason is recorded in the audit log. Nothing is lost, and when activity resumes the cases proceed in the order in which they stopped. Pausing the bookings process also suspends its cases that were in progress.
Where should you start?
Bookings are a good process to automate when requests come in every week and the team can already explain the rules. Before starting, it helps to have the following in place.
- The types of service, such as delivery, installation and technical visit, and how long each one takes.
- The calendars, one per team or per technician, and the criterion that decides which service goes to which calendar.
- The rules in writing, such as the areas covered, the minimum notice and the requests that should always go to a person.
- A mailbox for requests, so that they all arrive in the same place.
- The reply texts, for confirmations, changes, cancellations and the day-before reminder.
In an installation running in demonstration mode, processes can be tried out with sample emails, with no effect on the real systems.
The team does not lose control of the calendar. It spends its time on the requests that need a decision, not on those that resolve themselves. The bookings agent is built on the same foundation as the other agents, as explained in the article on what an AI agent for customer requests is, and other customer service automations are described in the communication and customer service solutions.
Frequently asked questions
What is a bookings agent?
It is an artificial intelligence agent that reads customers' written requests and proposes the next step, within the company's rules. The platform books the delivery, installation or technical visit in the right calendar and confirms it to the customer.
Does artificial intelligence choose the time slot?
In EngiAIFlow, no. AI reads the request. Calendar availability is worked out by the platform, without AI, according to the rules the team has set up.
What happens to a request that falls outside the rules?
It is not booked on its own. It reaches a person on the team with what has already been done and the decision still needed.
Does the agent also handle changes and cancellations?
Yes. Changes and cancellations requested by the customer are handled in the same way, and automatic notices go out when a booking is changed or cancelled.
Is it only for deliveries and installations?
No. It also covers technical visits and works in the same way for appointments at a clinic or services at a garage.
Can automated bookings be stopped at any time?
Yes. Pausing the bookings process also suspends its cases in progress. Whoever administers the installation can also stop, with a single command, all external actions and every incoming channel, with the reason recorded in the audit log. Nothing is lost, and cases proceed in the order in which they stopped.